I would argue that it doesn't say anything that conflicts with them. But you were citing old cases as authority for software being per se nonstatutory subject matter, and that's pretty clearly not true in the post-Bilski world.
Theoretically, if the only opinions that conflict with those old cases are from lower courts, then they're still valid precedent. However, you're quite right that in today's world you're unlikely to get a patent overturned because of Benson or Flook. This difference between theory and practice irks me.
Don't get me wrong, I don't think SCt Bilski does anything but muddy the water. But it's simply not the case that there is binding precedent that software is nonstatutory. That's all I was talking about.
From a practical standpoint, you're absolutely right, and I certainly hope nobody takes what I'm saying as legal advice.
From what's actually written in the Supreme Court opinions I've read, every software patent I've seen discussed on HN should be invalid. Benson and Flook were super clear. Diehr was long and hard to read, but when you boil it down, it didn't change anything; it merely clarified that software as part of an invention doesn't automatically make the whole invention nonstatutory. I haven't read Bilski, but it sounds like SCOTUS punted on clarifying things because they think it should be Congress's job.
Until Congress steps up and clarifies, I think the previous Supreme Court decisions should be law. But they aren't. I find that extremely annoying.
